Instrumentational operation and analytical methodology for ...

Atmos. Meas. Tech., 3, 1241–1254, 2010www.atmos-meas-tech.net/3/1241/2010/doi:10.5194/amt-3-1241-2010© Author(s) 2010. CC Attribution 3.0 License.

AtmosphericMeasurement

Techniques

Instrumentational operation and analytical methodology for thereconciliation of aerosol water uptake under sub- andsupersaturated conditions

N. Good1,*, H. Coe1, and G. McFiggans1

1Centre of Atmospheric Science, School of Earth Atmospheric and Environmental Sciences, University of Manchester,Manchester, UK* now at: Laboratoire de Meteorologie Physique, Blaise Pascal Univ., Clermont Ferrand, France

Received: 30 September 2009 – Published in Atmos. Meas. Tech. Discuss.: 1 February 2010Revised: 27 August 2010 – Accepted: 13 September 2010 – Published: 15 September 2010

Abstract. The methodology for the operation and calibrationof a hygroscopicity tandem differential mobility analyser(HTDMA) and cloud condensation nuclei counter (CCNc)for size-resolved measurements of aerosol water uptake arepresented. A state of the science aerosol thermodynamicmodel is used to benchmark the performance of the instru-ments. The performance, calibration and operation of theinstruments is then demonstrated in the field.

1 Introduction

The size and composition of aerosol particles determinestheir ability to take up water. The conditions they experiencein the atmosphere determine their behaviour. Atmosphericconditions and atmospheric aerosol can vary greatly spatiallyand temporally. Therefore understanding aerosol water-uptake in a range of conditions (including sub-saturated rel-ative humidities and the supersaturated conditions found inclouds) is necessary in order to accurately describe their rolein atmospheric physics and chemistry. In turn theoretical de-scriptions of aerosol water uptake can be informed by andmust be evaluated against observed behaviour.

In order for aerosol hygroscopic growth and CCN activ-ity to be characterised, and theoretical or empirical descrip-tions of this behaviour to be evaluated, these properties mustbe accurately measured. The hygroscopicity tandem differ-

Correspondence to:G. McFiggans([email protected])

ential mobility analyser (HTDMA) (Liu et al., 1978; Swi-etlicki et al., 2008) can be used to accurately determine sub-saturated hygroscopic water uptake of ambient aerosol parti-cle populations. These instruments require careful operationand calibration as no standard for their design has been de-veloped and applied consistently (Duplissy et al., 2009).

Various instruments have been developed to measure theability of aerosol particles to act as CCN (Nenes et al.,2001). One such instrument is the Droplet MeasurementTechnologies (DMT) CCN counter (CCNc) (Roberts andNenes, 2005). The DMT CCNc’s operation has been char-acterised in several studies (Rose et al., 2008; Lance et al.,2006). The DMT CCNc’s performance relies on the tempera-ture (T ) gradient set down a conductive column to generate aconstant centreline supersaturation to which the aerosol sam-ple is exposed.Lance et al.(2006) showed that modelling theCCNc’s supersaturation profile assuming theT measured onthe outside of the column is equal to theT directly on theinside of the wall is not adequate to determine the centrelinesupersaturation (S) precisely. The difference in temperatureresults from the thermal resistance of the columns’ walls.Lance et al.(2006) note that the thermal properties of thewalls may vary from instrument to instrument and with time.Rose et al.(2008) demonstrate the operation of a DMT CCNcunder a range of conditions.Rose et al.(2008) show that cal-culating the thermal efficiency (Lance et al., 2006) and cor-recting the supersaturation derived from a model of the in-strument, cannot always reconcile their measurements to the-oretical values from thermodynamic models. It is predictedtheoretically and shown in practice that the operating condi-tions (absoluteT , pressure and and flow rate) will change the

Published by Copernicus Publications on behalf of the European Geosciences Union.

http://creativecommons.org/licenses/by/3.0/

1242 N. Good et al.: Operation of HTDMA and CCN counter

centreline supersaturation set by a DMT CCNc for a giventemperature gradient (δT ) down the column (Roberts andNenes, 2005; Lance et al., 2006; Rose et al., 2008), howeverthe theoretical and inferred (from test aerosol calibrations)centreline supersaturation may not match.

In this paper a procedure to characterise and validatethe operation of a HTDMA and CCNc to measure atmo-spheric aerosol hygroscopic properties is outlined with theaim of achieving reliable and well characterised measure-ments of aerosol water uptake and potential CCN activity.The methodology for determination of operational supersat-uration is not reliant on accurate knowledge of the actualcentreline temperature profile. The centreline supersatura-tion is inferred from the observed CCN activity of aerosolof known composition with respect to a reference thermody-namic model. It is not derived from the temperatures mea-sured around the column. Only the repeatability of the tem-perature set-points between calibrations is assumed. Sucha procedure may be used to evaluate long-term instrumentdrift in actual temperature in response to set-point values andto compare with supersaturation control by nominal temper-ature control.

2 Instruments

2.1 HTDMA – principle of operation

A HTDMA operates by selecting a dry (<15% RH) quasi-monodisperse aerosol sample of diameter (D0) using a dif-ferential mobility analyser (DMA1). The aerosol sample isthen conditioned to a set relative humidity (RH) in a humidi-fier. The RH conditioned aerosol is then sized using a secondDMA (DMA2) conditioned to the same RH as the sample inorder to determine the equilibrium diameter (D) of the hu-midified aerosol. Data from a HTDMA is usually reported interms of the hygroscopic growth factor (GFD0,RH) which isdefined as:

GFD0,RH =D(RH)

D0(1)

The HTDMA operated here is described in detail byCubi-son et al.(2005) andDuplissy et al.(2009). In summary theHTDMA uses a gore-tex humidifier to condition the aerosolsample. The sample is held for a residence time of≈ 30 sec-onds prior to being sized by DMA2. DMA2 is operated bydiscretely stepping across a range of sizes i.e. as a differen-tial mobility particle sizer (Keady et al., 1983). DMA2 isstepped across a range of sizes to measure GFD0,RH in therange of 0.8 to 3 when operated at 90% RH, in order to coverthe range of GFs in the ambient environment. In this modeof operation the HTDMA takes≈10 min to step across therange of particle growth factors. DMAs 1 and 2 are bothoperated with volumetric sheath flows of 5.5 L min−1 andsample flows of 0.55 Lmin−1.

2.2 CCNc – principle of operation

The CCNc operated here is built by Droplet MeasurementTechnologies (DMT CCN model version 1) and its designis based on the CCNc instrument described byRoberts andNenes(2005). Rose et al.(2008) describe the DMT CCNversion 2, which is fundamentally the same instrument withsome minor differences. The CCNc exposes the aerosol sam-ple to a supersaturation as it is drawn along the centrelineof a column, separated from the walls by a near saturatedsheath air flow. The sample aerosol is conditioned to a con-stant temperature (TInlet) with respect to which all the othertemperatures in the CCNc are set. The temperature of thecolumn is controlled at the top (to a temperature (T1), mid-dle (T2) and bottom (T3), whereT3 > T2 > T1 ) in order toachieve a smooth temperature gradient down the walls of thecolumn. The walls of the column are continuously wetted.Heat and water vapour diffuses from the walls of the columntowards the columns centreline, whilst the sheath flow drawsthe diffusing gases down the length of the column. Watervapour diffuses quicker than air. Therefore a supersaturationdevelops as water molecules which are near saturation con-centration at point on the wall from which they originate, ex-perience lower temperatures as they diffuse towards the cen-treline. The temperature is lower because the air diffusingto the same point on the centreline originates from a pointfurther up the column which is cooler (because of theT gra-dient applied to the column). The temperature gradient mustbe small compared to the aerosol inlet temperature to main-tain a near constant centreline supersaturation.

Once particles reach the bottom of the column they enteran optical particle counter (OPC) located directly below thecolumn. The OPC signal is used to count the number of par-ticles in 20 bins between 0.5 and 20 microns in diameter. TheOPC data is recorded at 1Hz.

The CCNc is operated in 2 modes in order to determinethe size-resolved CCN activity of the sampled aerosol par-ticles. In the first mode, a DMA is used to supply a singledry size cut of aerosol, the CCNc then sets a series of differ-ent supersaturations. The number of particles which activateat each supersaturation is then used to investigate the CCNactivity of the aerosol. In the second mode of operation thesupersaturation is held constant whilst the DMA is operatedas a DMPS; stepping over a series of mobilities. The numberof particles which activate into CCN as a function of size canthen be used to investigate the CCN activity.

3 Experimental configuration

3.1 Sample charge equilibrium

Prior to entering the DMA, the sample aerosol is charge equi-librated using a Strontium-90 radioactive source. Aerosolscan exist in various charged states (e.g.Maricq, 2006; Vana

Atmos. Meas. Tech., 3, 1241–1254, 2010 www.atmos-meas-tech.net/3/1241/2010/

N. Good et al.: Operation of HTDMA and CCN counter 1243

et al., 2006). Sub-µm aerosol particles may be neutralor in positively or negatively charged states. The equilib-rium charging probability can be calculated (Fuchs, 1963;Wiedensohler, 1988), allowing the number of charges on par-ticles of a given mobility that are selected by a DMA to bedetermined if the raw (total number of particles detected bythe CPC at each mobility) number size distribution of theaerosol is known. The DMAs used to size select the aerosolin this study, select negatively charged particles.

When operating a HTDMA it is usually assumed that allthese particles sampled are singly charged. If the sample con-tains significant numbers of particles with multiple chargesthis may lead to erroneous interpretation of the data. The hy-groscopic growth of simultaneously sampled different sizedparticles cannot be differentiated analytically (Gysel et al.,2009). The larger multiply charged particles may also havedifferent growth factors due to the Kelvin effect and theirchemical composition, further increasing the measurementuncertainty. It is therefore recommended that the fraction ateach charge of the selected mobility is calculated in order tovalidate the HTDMA measurement (Duplissy et al., 2009).In the measurements presented here, multiply charged parti-cles usually make up less than≈20% of the number of par-ticles sampled at each size and as such the HTDMA data isinverted assuming all particles carry a single charge.

When operating a CCNc in tandem with a DMA, particleswith more than one charge may also be selected and sampledby the CCNc. The CCN activity of aerosols is sensitive totheir size (McFiggans et al., 2006) therefore the size of theparticles selected must be well defined. The DMA-CCNcmeasurement technique is often used in ambient and labo-ratory studies (e.g.Allan et al., 2008; Gunthe et al., 2009;Juranyi et al., 2009) and therefore the sampling of multiplycharged particles must be minimised or accounted for. Sev-eral methods for correcting for multiply charged particlesin DMA-CCNc measurements have been developed (Roseet al., 2008; Petters et al., 2007). The method used to correctfor multiply charged particles in the DMA-CCN measure-ments presented here is outlined in Sect.5.2.

The top panel of Fig.1 shows the fraction of particles ofa given charge that would be selected by a DMA as a func-tion of mobility across 3 example size distributions (from amarine environment, smog chamber experiment and a nebu-liser). The bottom panel shows the raw (averaged numberconcentration detected by the DMPS at each set mobilitywith no corrections applied) size distributions and the cor-rected (including multi-charge correction) number size dis-tributions (Williams et al., 2007). The number distributionsare plotted versus mobility diameter (Dp), which is definedas:

Dp (n = 1) =neC

3πµZp

(2)

wheren is the number of charges on the particle,e is theelectron charge,C is the Cunningham slip correction factor,

1.0

0.8

0.6

0.4

0.2

0.0Fra

ctio

n w

ith c

harg

e

2 4 6100

2 4 6

Mobility diameter (nm)

1

0

Nor

mal

ised

dN

dlog

Dp

1

05 6 7

1002 3

1

02 4 6 8

1002

Single charge Double charge Triple charge

raw corrected

Marine Smog chamber Nebuliser

Fig. 1. Example size distributions and the fraction of particles withsingle, double and triple negative charges at each mobility diameter.

µ is the viscosity of air inside the DMA andZp is the elec-trical mobility (Knutson and Whitby, 1975). For the marinedistribution there is an increase in multiply charged particlesbetween≈80 nm and≈200 nm. The largest contribution ofmultiply charged particles occurs forDps between the modesof the distribution because of the higher number of largerparticles. For the smog chamber distribution there are moremultiply charged particles at sizes below the mode than sin-glets, above the mode the singly charged fraction is close to1. The number size distribution generated by the the nebu-liser is composed mostly of singly charged particles (>90%).The fraction of multiply charged particles increases with sizeup to≈10% atDp=200 nm.

For the nebuliser distribution the influence of multiplycharged particles is generally small and will not impact mea-surements significantly. For the smog chamber distribution,the effect of multiple charging can be large as noted by otherauthors (King et al., 2009; Duplissy et al., 2009). For the ma-rine distribution we see that there are some diameters whichhave quite large contributions, e.g. around 100 nm, whilstothers do not. It is therefore necessary to have a procedure tocorrect this data for multiply charging.

3.2 HTDMA configuration

Duplissy et al.(2009) review the instrumental parametersimportant to maintaining reliable HTDMA operation anddata collection. The flows and temperatures in the DMAsare monitored continuously as described by (Cubison et al.,2005) to ensure they operate reliably and stably. An impor-tant feature of the HTDMA set-up to note is that the tem-perature is measured inside the bottom of DMA2 and in thesheath excess at the top of DMA2, to ensure the RH is con-stant and well defined in the DMA. The residence time of

www.atmos-meas-tech.net/3/1241/2010/ Atmos. Meas. Tech., 3, 1241–1254, 2010


��

��

��

��

��

��

� ��

��

��

��

� ��

� ��

� ��

! ��

" ��# $�$�

��

�� %��$��% �� &��

Fig. 2. CCN-DMA experimental set-up. Schematic showing howthe DMA and CCNc are coupled. The 4 way-valve can be set in2 positions, such that the CCN is either sampling aerosol from theDMA or directly from the sample and the DMA is either supplyingthe CCNc or the CPC. For the calibrations the CPC is connectedso it samples simultaneously to the CCNc and the make-up flow isreversed to main 1 L min−1 sample flow.

the humidified aerosol sample prior to sizing in DMA2 is≈30 s, so that the sample has time to equilibrate (Sjogrenet al., 2007). The accuracy the HTDMA’s measurements areensured by calibrating as described in the proceeding Sect.4.

3.3 CCN configuration

Size-resolved measurements of CCN activity are performedby coupling the CCNc to a DMA. Figure2 illustrates theexperimental set-up. If the sample aerosol is not dry it isdried using a diffusion drier (e.g. Perma Pure, MD-110).The sample is then charge equilibrated using a Strontium-90 or Polonium-210 radioactive source (depending on sourceavailability in a particular project). A quasi-monodispersesize cut is then selected using a DMA. The DMA is operatedwith a sheath flow of 10 litres per minute and a sample flowof 1 litre per minute. The DMA’s sheath flow is dried us-ing a Nafion (Perma Pure, PD-200T-12) drier. The RH andT of the sample, sheath and sheath excess flows are mon-itored continuously using capacitive probes (Rotronic, Hy-groClip) (Williams et al., 2007). The size selected sample isthen either split between the CCNc and a condensation par-ticle counter (CPC, TSI 3010) or to either one of the CPC orCCNc depending on the chosen configuration.

For calibrations the aerosol is sampled by both the CPCand the CCNc. The sample flow to the CPC is diluted with0.5 litres per minute of particle free air, to maintain a flowof 1 litre per minute through the DMA. The sample flow iscontinuously monitored using a differential pressure gauge.The length of the sample lines to the top of the column of theCCNc the CPC’s inlet are matched so that any differences

in particle loss are minimised. For ambient measurements, aconfiguration where either the CPC or the CCN sampled thesize selected aerosol is frequently used. An automated 4-wayvalve switches between the CCNc and CPC every hour. Thisallows measurements of the total CCN number concentrationduring the hour when it is not connected to the DMA and theconfiguration then allows measurements of the size-resolvedCCN number (NCCN(S,D0)) and CN number (NCN(D0)) tobe compared, as well as the total CCN concentration. Alter-natively the configuration is left the same as for the calibra-tions and simultaneousNCCN(S,D0) andNCN(D0) distribu-tions are measured. The CCNc is operated throughout the ex-periments described here with a total flow rate of 5 Lmin−1,with a 10:1 sheath to sample flow ratio. The CCNc’s columnwas wetted at a rate of≈8 m L min−1 (the medium flow set-ting). T1 is set to 1 K aboveTinlet, the OPC (optical particlecounter)T is set to 3 K aboveT3.

4 Instrument calibration

Prior to experiments the performance of the instruments isalways characterised and validated. In the case of the HT-DMA this means calibrating its DMAs, capacitive RH andT probes and sampling 2 nebulised aerosols of known com-position and growth factor. In the case of the CCNc thismeans first calibrating the DMA used to size select the sam-ple aerosol, calibrating the CCNc’s sample and sheath flowsand then calibrating the supersaturation using 2 nebulisedaerosols of known composition.

4.1 DMA calibration

Both the HTDMA and CCNc rely on DMAs to size select theaerosol sample prior to determination of their hygroscopicgrowth and CCN activity. It is therefore important to verifythat they are sizing correctly. The DMA’s sheath flow metersare calibrated using a primary air flow standard (Sensidyne,Gilibrator) and the high voltage supplies are tested to ensurethe correct voltage is being set.

The DMAs used to size select the aerosol samples are thencalibrated using latex spheres (Duke Scientific). The latexspheres are nebulised and sampled by the DMAs operatingin DMPS mode, with over-sampling around the peak. TheHTDMA’s DMA1 is connected directly to the CPC, whilstthe output from the DMA to the CCNc is switched so thatonly the CPC draws the sample flow (as shown in Fig.2), bydoing this the same operating conditions (most importantlythe sheath and aerosol flow rates) are maintained. The dis-tributions are then fitted to determine the size at the peak ofthe distribution. This is repeated for a series of sizes between92.5 nm and 598 nm. A linear regression is then used relatethe diameter set by the DMA to the PSL diameter to pro-duce a correction to the DMA sizing (Williams et al., 2007).One of the limitations of these calibrations is that nebulising



Table 1. DMA PSL calibrations for the DMA supplying the CCNc and the HTDMA’s first DMA. The fit error is the difference between theapplied linear fit and the measurement.

PSL (nm) PSL Std. (nm) DMACCNc (nm) Fit Error (nm) DMAHTDMA (nm) Fit Error (nm)

92 7.0 94.4 1.91 92.5 0.59199 3.4 197 1.30 201 0.91299 4.1 296 1.50 303 2.91596 6.5 592 0.91 594 1.91

spheres smaller than≈90 nm is not possible using our stan-dard techniques (Cubison et al., 2005). Therefore calibra-tions are carried out in the laboratory using an electrosprayaerosol generator (TSI, 3480) to produce sub-90 nm particlesof known size for calibration. The electrospray generator canbe used to produce particles down to 4 nm (Chen et al., 1995).Therefore the full size range of the DMAs can be calibratedfor in the laboratory prior to deployment in the field ensuringthe DMA is operating consistently. Table1 shows the diame-ters set by the CCNc’s and the HTDMA’s first DMA and thecorresponding correct PSL diameter. From these calibrationsthe resulting difference in calibrated versus un-calibrated di-ameter was up to 4 nm.

4.2 HTDMA calibration and validation

The capacitive RH and temperature sensors used to moni-tor the conditions in the instrument (Cubison et al., 2005)are calibrated against a dew point hygrometer (the same dewpoint hygrometer used to measure the RH in DMA2) to en-sure consistent measurements. The dew point hygrometeris serviced periodically off site to ensure its accuracy. Dryscans are used to calibrate any offset between DMA1 andDMA2 (Gysel et al., 2009; Duplissy et al., 2009). Dry scansare performed with roughly weekly frequency during fieldprojects to ensure the continued performance of the HTDMAand also to define the width of the HTDMA’s transfer func-tion (Gysel et al., 2009). For this project the offset betweenDMA1 and DMA2 was quite large in the range of 9% to11% across the range of selected dry diameters. This demon-strates the importance of validating DMA offset and well asthe absolute sizing of the first DMA as described byGyselet al.(2009). The operation of the HTDMA was verified us-ing ammonium sulphate ((NH4)2SO4) and sodium chloride(NaCl) test aerosols. The test aerosol ((NH4)2SO4 and NaClin turn) is nebulised (using a TOPAS atomiser). The nebu-lised aerosol is then dried (to<15%RH), charge equilibratedand sampled by the HTDMA. The HTDMA then performs ahumidogram, whereby the RH of the aerosol is set to a seriesof values covering the operational range of the instrumentand the growth factor is measured. A dry size of 150 nmis selected for the calibrations because, above≈100 nm, theKelvin term is close to 1 (McFiggans et al., 2006) thus reduc-

2.6

2.4

2.2

2.0

1.8

1.6

1.4

1.2

1.0

GF

150m

,RH

90858075706560RH (%)

NaCl (NH4)2SO4

Fig. 3. HTDMA ammonium sulphate and sodium chloride calibra-tions made during the MAP campaign.

ing artefacts resulting from the sizing uncertainty ensuringstraightforward comparison of the data with other measure-ments and models. The HTDMA used here (Cubison et al.,2005) has the ability to measure effloresced aerosol water up-take. The dry size selected aerosol can be humidified up tothe set point RH or it can be passed through a pre-humidifierwhich increases the RH of the sample to> 90% before it isreduced to the set point RH in the humidifier. Figure3 showsthese calibrations for the Marine Aerosol Production (MAP)project. From Fig.3 it can be seen that the agreement be-tween the measured and theoretical growth factors is withinthe measurement error. The measurement error results fromthe stability of the TDMA estimated to be± 0.02 in growthfactor space for a well calibrated TDMA (Duplissy et al.,2009) and the uncertainty of the RH measurement (± 1.5%for the dew point hygrometor). As a result of the operatingprocedure described the calibration measurements fall withinthese limits.



4.3 CCNc calibration

The CCNc’s sample and sheath flows are calibrated using aprimary flow standard (Sensidyne, Gilibrator). The sizingof the OPC is verified by sampling nebulised 1000 nm la-tex spheres (Duke Scientific). The CCNc is calibrated usingammonium sulphate and sodium chloride test aerosol gen-erated in the same manner as for the HTDMA calibrations(described in Sect3). The CCNc is set-up as described inSect. 3.3. For the calibrations a single size is selected bythe DMA and the CCN concentration is measured at a seriesof supersaturations, simultaneously the total particle numberconcentration at that size is measured by the CPC. The CCNand CN concentrations are then used to calculate the acti-vated fraction (AF(S,D0)) as defined in Eq.3:

FA(S,D0) =NCCN(S,D0)

NCN(D0)(3)

whereNCCN(S,D0) is the CCN number concentration at theset supersaturation (S) andD0 andNCN(D0) is the numberconcentration of condensation nuclei (CN) at the selectedD0. The FA is then used to calculate the critical supersat-uration (Sc) of the particles at the selectedD0. Sc(D0) isdefined as the supersaturation at which FA(S,D0) is equal to0.5. This definition ofSc(D0) relies on several assumptionswhich are discussed in Sect.5.3. In order to defineSc asaccurately as possible theS at which FA = 0.5 needs to befound. This is achieved by incrementally increasing the set-point S such that FA increases from 0 to 1. In order that FA= 0.5 is well defined the resolution of the set-pointS is in-creased for FA values around 0.5. Increments in the set-pointS of 0.02% are used (equivalent to steps inδT of ≈ 0.02 K).An S should be set such that FA=0, this helps verify there areno leaks. It is also important that the maximum FA is mea-sured. If the CPC and OPC are in agreement it should be 1,however there may be a systematic offset resulting from theircalibration. Once all reasons for a discrepancy have been re-moved (e.g. leaks and flow imbalances are perhaps the mostlikely causes) a correction factor may be applied to the data.The accuracy of most commercial CPCs and OPCs is usually± 10% so a correction factor to the FA is usually no morethan 15%. For the example experiments shown here a cor-rection factor of +4.5% is applied to the measuredNCCN.

It is important define which particles detected by theCCNc are activated droplets. The CCN counter measuresthe size of the particles exiting the supersaturated column us-ing an OPC. In this paper particles of 1µm and above areconsidered as activated CCN.Lance et al.(2006) suggest amethod for determining which droplets sizes in the OPC areCCN, which may offer an improvement. It is expected thatparticles in the size range sampled will be activated dropletsif they are 1µm or greater in diameter when detected in theOPC. This assumption will likely be tested at the lowestS setpoint (0.08%) (Lance et al., 2006). An empirical approach todetermine the bin to choose may improve on the method pre-

sented here but without knowing the growth kinetics of thesampled aerosol particles it is hard to say if the effect willbe large, we assume it will not be for the ambient aerosol.In the case of the calibration aerosol the critical diameter ispredicted by the thermodynamic model (ADDEM,Toppinget al. (2005a)) as part of the critical supersaturation calcu-lation. The predicted critical diameters for the calibrationaerosols are consistent with using 1µm as the cut-off for ac-tivated particles down to the lowest set supersaturation.

Figure 4 shows ammonium sulphate calibrations per-formed during the MAP project. 5D0s were selected (30,40, 50, 70 and 100 nm) in order to calibrate across the rangeof S set by the CCNc during the project. The settling time ofeachS depends on the size of the temperature change andthe direction (downwards steps inT take longer than up-wards steps). Settling times of between 5 and 1 minutes wererequired, therefore each supersaturation set-point was heldfor 4 to 10 min to give adequate settling and sampling time.The settling was verified by making sure the CCN numberhad stabilised as well as the temperature readings from theCCNc’s diagnostics. The total number concentration is keptbelow 3000 particles per cm3 (by monitoring the CPC mea-surement) to avoid coincidence errors in the CCNc’s OPC.NCCN andNCN at each set point are averaged, removing dataduring the settling period and then used to calculate the ac-tivated fraction. A series of supersaturations are set in theCCN counter to build up an activation spectrum from 0 to1 as described above. A sigmoid function is then fitted tothe data from which theS at which FA = 0.5 is determined.Figure4 shows the calibrations and fitted sigmoids for am-monium sulphate calibrations performed during the MAPproject. Panels a to e in Fig.4 show the fitted sigmoids(black lines) and with the correction factor of 1.045 applied(grey dashed lines), the correction makes a negligible differ-ence to the derivedSc. The same procedure is then repeatedfor the sodium chloride test aerosol, using the same mobilitydiameters. The mobility diameter calculated by the DMA’ssoftware assumes the particles are spherical, however for ex-ample sodium chloride particles tend to be cubic. Constantshape correction factors of 1.02 and 1.08 (Rose et al., 2008)are therefore applied to the ammonium sulphate and sodiumchloride calibration particles respectively to correct for this.

The results are then compared to their theoreticalSc (Top-ping et al., 2005a,b). Panel f in Fig.4 shows theSc derivedfrom the set-pointS compared to the theoretical values ateachD0 for the ammonium sulphate calibration. The sameprocedure is repeated for the sodium chloride calibrations.The calibrations using the different salts are then compared,if the instrument is operating properly then they should fallon the same line when plotted as set-point versus theoreti-cal supersaturation. Figure5 shows the sodium chloride andammonium sulphate calibrations for the MAP project plot-ted together. Figure5 illustrates the improvement in the de-termination of CCNc’s centreline superaturation comparedto the factory settings. The five supersaturation set-points



1.0

0.8

0.6

0.4

0.2

0.0

FA

1.00.80.6S Set

1.0

0.8

0.6

0.4

0.2

0.0

FA

0.20.1S Set

1.0

0.8

0.6

0.4

0.2

0.0

FA

0.30.20.1S Set

1.0

0.8

0.6

0.4

0.2

0.0

FA

0.40.30.2S Set

1.0

0.8

0.6

0.4

0.2

0.0F

A0.60.50.40.3

S Set

1.0

0.8

0.6

0.4

0.2

0.0S

c S

et1.00.80.60.40.20.0

Sc ADDEM

a b

d

c

e f

1:1

Fig. 4. CCNc ammonium sulphate calibrations. Panels a-e showAF vs.Sset for 30, 40, 50, 70 and 100 nmD0 particles made duringthe MAP campaign. Panel f shows the set pointS vs. theory.

used during the MAP campaign were 0.07, 0.15, 0.3, 0.5 and1.0%. The post-calibration values were 0.08, 0.18, 0.37, 0.62and 1.28% respectively,differences of between 0.01 and 0.28%.

4.4 Theory

It is only possible to accurately calibrate the CCN countersupersaturations and validate the operation of the HTDMAusing a suitably accurate thermodynamic model with knowntest aerosol. The state of the science Aerosol Diameter De-pendent Equilibrium Model (ADDEM) as described undersubsaturated conditions inTopping et al.(2005a) and for su-persaturated conditions inRissman et al.(2007) is used. As-suming the inorganic salt test aerosol reaches its equilibriumsize in the HTDMA and activated droplets reach their crit-ical diameter in the CCNc, ADDEM can accurately predictthe CCN activity and hygroscopic growth factors of the cal-ibration aerosol within the instrument resolutions. Becausethe CCNc uses a temperature gradient to generate a super-saturation, choosing a temperature to run the model at is notstraightforward. In general, for calibrations, the model pre-dictions at 298.15 K were used. A more detailed study usinga model of the droplet growth inside the CCNc might be ableto back out a slightly more accurate calibration, by calculat-ing the exact point in the CCNc where the droplet activated.However this would rely on knowledge of the absolute tem-perature inside the column and would be inconsistent withour approach. The variability inSc across theT investigatedis sufficiently small to ignore (± 5% at 1%S reducing to±1% at 0.01%S) without biasing the results significantly.

1.0

0.8

0.6

0.4

0.2

0.0

Sc

Set

1.00.80.60.40.20.0

Sc Theory

(NH4)2SO4

NaCl

Linear fit (R2= 0.992)

1:1 line

Fig. 5. CCN ammonium sulphate and sodium chloride calibrationsfrom the MAP campaign. Set point vs. theoreticalSc.

5 Data analysis

5.1 HTDMA data analysis

A detailed description of the HTDMA data analysis proce-dure is described by (Gysel et al., 2009) so only a brief out-line is given here. Dry scans are performed regularly (at leastonce a week during continuous operation and at the start andend of the measurement period), so the DMA offset and thesystem transfer function can be corrected for. The raw datais then processed according to the procedure and inversionroutine described byGysel et al.(2009).

5.2 DMPS-CCNc data analysis

In the DMPS-CCN mode of operation the DMPS is used tosupply the CCN counter with an aerosol sample at a series ofmobilities. This enables CCN number size distributions to begenerated. The CCN distributions are then used to calculateFA(S,D0). The data are obtained and analysed as follows:

5.2.1 Raw data

The aerosol is sampled by the CCN and CN counters as de-scribed in Sect.3. This generates raw CCN and CN numbersize distributions. The first 6 s of data at each mobility is dis-carded to allow the instruments to settle. The remaining 18 sof data at each diameter is then averaged. Figure6 shows anexample of raw and averaged raw CCN and CN number sizedistributions for nebulised sodium chloride aerosol.

5.2.2 Multi-charge correction

The raw distributions are corrected for the effects of multiplycharged particles as follows. Throughout these calculations



��

��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

Fig. 6. Example DMPS-CCN distribution of nebulised sodiumchloride aerosol at 0.26% supersaturation. The blue squares showthe averagedNCN, the orange squares show the averagedNCCN.The crosses indicate the mobility diameter. The dashed lines showthe raw data.

the DMA transfer function is assumed to be ideal (Alofs andBalakumar, 1982) that is; triangular with a half width of:

δZ =Qa

2πLVln(r) (4)

whereδZ is the half width,Qa is the sheath flow,L isthe length of the DMA’s central rod,V is the voltage appliedto the DMA’s anode andr is the ratio of DMAs outer cylin-der inner radius to the radius of the central rod. The preciseshape of the transfer function is not important for the pro-ceeding data analysis provided it is symmetrical. The AFis used to deriveSc, so diffusional broadening for examplewill be cancelled out whenNCCN(S,D0) andNCN(D0) aredivided.

The charging probabilities at each selected mobility diam-eter are calculated (Wiedensohler, 1988). The assumption isthen made that particles detected by the CPC at the lowestmobility are all singly charged. This appears to be a goodassumption in the studies presented here as closed size dis-tributions are observed i.e. at the largest diameter the num-ber tends to zero. As noted byPetters et al.(2007) this doesmean that CCN measurements are required up to lower mo-bilities than might otherwise be of interest. The number ofparticles detected at the smallest mobility is then divided byits single charge probability to calculate the actual particlenumber. This assumption is repeated for increasing mobili-ties until their mobility is at least twice that of the lowest mo-bility (Zp min). At this point doubly charged particles can beaccounted for and similarly, for triply charged particles oncethe particle mobility is 3 times the smallest mobility. Thenumber of doubly and triply charged particles are subtractedfrom the total number measured (N ′) at each mobility to give

��

��

��

��

��

��

�

��

� �

��

Fig. 7. Example plot of FA vs.S, with a sigmoid fit used to deter-mineSc measured during the MAP campaign.

the number of singly charged particles. Once the number ofsingly charged particles is known, the total number is calcu-lated by dividing by the single charging probability of thatdiameter. The multi-charge correction can be described ac-cording to:

NS(Zp) =N(Zp)

εS(Zp)

ND(Zp) =N(Zp/2)

εD(Zp/2)

NT(Zp) =N(Zp/3)

εT (Zp/3)

when:Zp <Zpmin

2

N(Zp) =N ′(Zp)

εS(Zp)

when:Zpmin

3< Zp ≤

Zpmin

2

N(Zp) =N ′(Zp)−ND

εS(Zp)

when:Zp ≥Zpmin

3

N(Zp) =N ′(Zp)−ND −NT

εS(Zp)

whereNS is the number of singly charged particles,ND isthe number of doubly charged particles,NT is the number oftriply charged particles,εS is the single charging probabil-ity, εD is the double charging probability andεT is the triplecharging probability. Both the CCN and CN number sizedistributions can be calculated using this method.



��

��

��

��

��

��

�

� � � � � �

��

��

��

Fig. 8. Example plot of FA vs.D0, with a sigmoid fit used to deter-mineD50 measured during the MAP campaign at 0.37% supersat-uration.

5.2.3 DerivingSc

A method for calculating the critical supersaturation from themeasured CCN and CN number size distributions is now de-scribed. The errors associated with each of these steps andthere propagation is described in detail byIrwin et al.(2010).Firstly the charge corrected CCN and CN number size distri-bution to calculate FA(S,D0). The activated fraction is thenused to deriveSc. In this analysis it is assumed that whenFA(S,D0) equals 0.5 thenS is equal toSc, i.e. all the particleswith diameters up toD0 under the DMA transfer function areactivated CCN at thisS. The work presented here uses twomethods to calculateSc using the CCN and CN number sizedistributions from the DMPS-CCN set-up. One method is toplot FA(S,D0) at each of the supersaturations set by the CCNcounter. A sigmoid fit function is then applied to data to de-termine the supersaturation at which 50% of the particles areactivated, which has been defined as theSc. This method hasthe advantage that theD0 is held constant, but the disadvan-tage that only a limited number of supersaturations (5) areavailable to fit the data. Figure7 shows an example of sucha fit for the data measured during the MAP cruise. Alterna-tively a sigmoid fit of activated fraction vs. diameter can bemade. The fit function is then used to determine theD50, thediameter at which 50% of the particles activate at a given su-persaturation setting. Thus FA(S,D0 = D50)=0.5 andS=Sc

for D0=D50. The D50 method has the advantage that it isconstrained by more data points than the supersaturation fit,howeverD0 is variable so the behaviour at a constant diam-eter is not captured directly. Figure8 shows an example of aD50 derivation for the data measured during the MAP cruise.For the MAP project FA vs.S plots are used to derivedSc, sothatD0 remains constant. TheD50 method is applied only tocheck data consistency.

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Fig. 9. Ideal mobility and diameter transfer functions for 30, 100 &300nm particles. The vertical lines illustrate the difference betweenthe selected mobility diameter and the mobility diameter at 50% ofthe area of the transfer function.

5.3 DefiningSc

The definition ofSc used in these analyses relies on assump-tions about the system’s transfer function, (the penetrationprobability of a particle of a given size through the instru-ment) and the mixing state of the sample aerosol.

5.3.1 Transfer function

It is assumed that the system’s particle mobility transfer func-tion is symmetrical and that the aerosol is internally mixedwith respect to its CCN activity i.e. the particle penetra-tion efficiency through the DMA, CCN counter and con-necting tubing is symmetrical about the mobility set by theDMA. The ideal mobility diameter DMA transfer functionis slightly asymmetrical. However this translates to a verysmall difference in the actual diameter at 50% activation(Rose et al., 2008), but can be taken into account for com-pleteness (Fig.9 illustrates the correction required for 30,100 and 300 nm particles, the range of sizes typically sam-pled). Size dependent losses in the system could also cause abias in the transfer function, this is minimised by maintaininga sheath flow to sample flow ratio of 10:1, ensuring the widthof the transfer function is narrow enough that losses will notbe significantly different across its width.

5.3.2 Mixing state

The mixing state of the aerosol is a potentially important fac-tor. If the aerosol is externally mixed then the supersatura-tion at AF equals 0.5 may not represent theSc well. Forexample if there are 2 distinct particle hygroscopicities ofequal number fraction when FA = 0.5 the supersaturation willfall between their actualSc. It is therefore important to ver-ify that the sample is suitable for the analysis methods de-scribed. One way to do this is to examine the hygroscopicgrowth factor distributions for the aerosol being sampled. Ifthe growth factor distributions show that the aerosol is largelyinternally mixed i.e. the particles at anyone size are of the



��

��

��

��

��

��

��

��

��

��

��

��

��

��

Fig. 10. The Celtic Explorer cruise track between 26th June 2006and 5th July 2006 during the MAP campaign.

same composition thenSc can be calculated using the meth-ods described. As CCN activity is more sensitive to the parti-cle dry diameter than sub-saturated hygroscopic growth, thisassumption should hold, e.g. it is unlikely an aerosol whichappears internally mixed at 90% RH will have widely vary-ing CCN behaviour at a single dry size. If the growth factordistributions show that the aerosol is externally mixed then adifferent analysis approach must be adopted.

6 Example of operation in the field

The combination of HTDMA and CCNc was deployed in thefield as described in this paper on the Celtic Explorer duringthe MAP campaign for 2 weeks in June and July 2006. Themarine environment is a source of primary and secondaryaerosol which significantly impacts upon the global radiativebudget. The potential role of biological aerosol sources inthe marine environment has been highlighted (O’Dowd et al.,2004). Therefore a quantitative understanding of the mecha-nisms via which marine aerosol is formed and its propertiesis required. The Celtic Explorer provided a ship-borne plat-form for the in-situ study of Marine Aerosol in and aroundbiologically active waters. Figure10 shows the ship’s cruisetrack. The area surveyed was off the West coast of Ireland inthe North-East Atlantic Ocean. The air masses sampled wereof Marine origin from over the open ocean as well as someperiods where the air had passed over continental Ireland andcoastal areas.

Figure11shows the number size distribution measured us-ing a DMPS (Williams et al., 2007). The number size distri-bution is mostly dominated by particles in the Aitkin and Ac-cumulation size ranges. When the total (integrated) numberconcentration is less than≈500 cm−3 the number of Aitkin

�

�

��

�

�

�

��

�

�

�

��

��

��

��

��

��

��

Fig. 11. The aerosol particle number size distribution measuredduring the MAP Celtic explorer campaign.

and Accumulation sized particles is similar and the resultingbi-modal spectrum is typical of a clean marine environment(Hoppel et al., 1986). Frequently the air being sampled ischaracterised by higher number concentrations of the orderof ≈103 cm−3. In comparison to the “clean” conditions theconcentration in the Aitkin mode is most numerous. Onlyoccasionally are significant numbers of sub-10nm particlesobserved, during more polluted episodes.

Measurements of the hygroscopic growth factor were per-formed at 90% RH forD0s of 30, 50, 100, 150 and 200 nm.The size-resolved CCN activation was measured at 5 setsupersaturations forD0s between 30 and 300 nm. The 5calibrated supersaturations were 0.08, 0.18, 0.37, 0.62 and1.26%.

6.1 HTDMA measurements

Figure12 shows the time series of the hygroscopic growthfactor distributions retrieved from the HTDMA averagedover 4 h intervals. The grey areas on Fig.12 indicate pe-riods where the HTDMA was being calibrated. The distri-bution of hygroscopic growth factors at eachD0s showedlittle evidence of external mixing with respect to water up-take, with most of the humidified aerosol contained withina single mode. Exceptionally, a second relatively hydropho-bic mode appears, most likely attributable to localized pri-mary emissions from passing ships. The mean hygroscopicgrowth factor of the dominant mode varied between 1.3 and1.8; particles with smaller dry diameters showing slightlylower growth factors. The small decrease in the growth factorof the aerosol with decreasingD0s which can be attributed tothe Kelvin effect indicates that the composition of the aerosolmay be relatively homogeneous across the measured dry di-ameter range.



��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

� � �

� �

Fig. 12. Time series of the hygroscopic growth factor distributions measured during the MAP campaign, at the 5 setD0 in panels a to e.

�

�

��

�

��

��

��

��

��

��

��

�

�

��

�

��

��

��

��

Fig. 13. Time series of FA(S,D0) at each of the 5 supersaturations during the MAP project.

The observation of a more hygroscopic mode (GFD0,90of ≈1.5–1.8) persisting throughout the measurement periodis common to most HTDMA observations made in the ma-rine environment (Swietlicki et al., 2008) including the NorthEast Atlantic (Swietlicki et al., 2000). The growth factorof the more hygroscopic mode is consistent with an aerosolthe composition of which is dominated by sulphate. For ex-ample a pure 100 nm diameter ammonium sulphate aerosolparticle has a growth factor of 1.70 at 90% RH. However,the resolution of the HTDMA limits exact determination ofthe aerosol composition. Internal mixtures of more hygro-scopic solutes for example sodium chloride or ammoniumbisulphate combined with less hygroscopic compounds suchas organics could result in growth factors in the more hygro-scopic range.

A very hygroscopic growth factor mode (GFD0,90>≈1.8)is very rarely observed. A very hygroscopic mode is associ-ated with sea-salt (sodium chloride) aerosol. This indicatesthat there was not a large amount of sea-salt aerosol beingproduced. It may also be that sea spray aerosol is processedinto an internal mixture or other compounds (Svenningsson,2007) which cannot be resolved by the HTDMA.

6.2 CCNc measurements

Figure13shows FA(S,D0) measured at the 5 set supersatura-tions. FA(S,D0) is used to calculateSc using the two meth-ods outlined previously; by fitting FA(S,D0) againstD0 toderiveD50 and henceSc, and by fitting FA(S,D0) against the5 set supersaturations to deriveSc. Figure14shows the time



Table 2. Averageκ values and standard deviations (std.) from theproject, computed assuming the surface tension is that of pure water.D50 denotes FA vs.D0 fits andS Scan denotes FA vs.S fits.

Method κ (S,D0) κ κ κ

D50 0.83(0.08%) 0.82(0.18%) 0.85(0.37%) 0.79(0.62%)std. 0.33 0.28 0.49 0.45S Scan 0.89(30 nm) 0.80(50 nm) 0.76(100 m) 0.83(150 nm)std. 0.35 0.4 0.37 0.43HTDMA 0.40 0.33 0.31 0.32std. 0.09 0.06 0.05 0.05

series ofD50 at the 5 set supersaturations. Figure15 showsthe time series ofSc derived by fitting FA(S,D0) againstS.The CCN data is averaged onto a the same 4 h time base asthe HTDMA’s data. The derivedScs generally fall betweenthose of ammonium sulphate and sodium chloride. The blackboxes in the Figs.14and15highlight data points which havebeen fitted outside the operational range of the instrument.When theD50 is below the smallest size set by the DMA(30 nm) or theSc from the fitted sigmoid is substantially be-low the lowestS set by the CCN counter.

In order to ensure that the CCNc’s calibrations remainvalid the operational conditions during the experiment werekept are close to the calibration conditions as possible. Theinlet temperature varied from 296.4 K to 299.5 K (except fora 5 h period when it dropped to 295.3 K) and the sample pres-sure varied from 1001 mb to 1026 mb. Therefore accordingto the calculations of (Roberts and Nenes, 2005) the changein centreline supersaturation for these variations will be small≈0.01% and will not bias the measurements. A compar-ison between theSc values derived from the two methodsdescribed in this paper can be made by comparing theκ pa-rameters (Petters and Kreidenweis, 2007) derived from thedifferent values ofSc. Table2 shows theκ values from thetwo methods of derivingSc. The κ values for the 2 meth-ods of derivingSc are consistent, covering a similar range ofvalue.

The hygroscopic growth factors were also used to calcu-lateκ. κ derived from the DMA-CCN can be compared withthat from the HTDMA to examine the relationship betweenhygroscopic growth and CCN activity. Comparing theκ val-ues derived from the HTDMA and the CCNc we see from Ta-ble2 that on average theκ values from the CCNc are higher.There is also more variability in the CCNc derivedκ val-ues. The difference in the derived values may be due to aninstrument bias, however as demonstrated a rigorous opera-tional and calibration procedure was followed. The calibra-tion aerosol used for the CCNc covered the operational su-persaturation range and produced consistent results through-out the campaign. The same test aerosol measured by theHTMDA produced results consistent with theory. It may betherefore that it is the composition of the aerosol causing the

2

3

4

567

100

2

D50

(nm

)

26/06/06 28/06/06 30/06/06 02/07/06 04/07/06

Date

0.08% S 0.18% S 0.37% S '0.62% S

ADDEM (NH4)2SO4 ADDEM NaCl

Fig. 14. Time series ofD50 at 0.08, 0.18, 0.37 and 0.62 %S mea-sured during the MAP campaign. The dashed lines show the theo-reticalD50 at each of the supersaturations for ammonium sulphateand sodium chloride.

4

6

0.1

2

4

6

1S

c (%

)

26/06/06 28/06/06 30/06/06 02/07/06 04/07/06Date

D0=30nm D0=50nm D0=100nm D0=150nm

ADDEM (NH4)2SO4 ADDEM NaCl

Fig. 15. Time series ofSc at 30, 50, 100 and 150nm measuredduring the MAP campaign. The dashed lines show the theoreticalSc

at each of the supersaturations for ammonium sulphate and sodiumchloride.

difference. Several properties of the aerosol could explainthe difference. The non-idealality of the solution (i.e. activ-ity coefficient) might change significantly between 90% RHand the point of activation. The Kelvin term may not be ac-curately represented assuming the surface tension is equal tothat of water. The increase in the importance of the Kelvinterm at higher RHs means that if there are surface active com-pounds in the aerosol a reduction in the surface tension mayresult in an apparent increase inκ at the point of activation.Sparingly soluble compounds in the aerosol particles couldalso explain the differences. Solutes which are solid at 90%RH but which deliquesce between 90% and the point of CCNactivation would lead to an increase in the derivedκ. Furtherwork is needed to fully investigate the significance of the re-sults beyond the scope of this paper.



7 Discussion and conclusions

We have demonstrated a robust method of repeatedly cali-brating a field deployable system of CCN and HTDMA todetermine the size-resolved hygroscopic growth and CCNactivity of atmospheric aerosols. The instruments are shownto be accurately calibrated and their performance is verified.The operation of the instruments is then demonstrated in thefield where the properties of the aerosol sampled are notknown a-priori and the data analysis procedures (includingcorrection factors) are described.

Accurate calibrations are vital to ensure reliable measure-ment of hygroscopic growth and CCN activity. The perfor-mance of the HTDMA is validated with measurements ofammonium sulphate and sodium chloride test aerosols. TheRH and the growth factor measured by the HTDMA agreewith the state-of-the-science theoretical description withinmeasurement uncertainty. Thus operation of the HTDMA isverified and no RH correction factor is required. The CCNcrequires calibrating in order to determine theS to which theaerosol sample is exposed to. TheS at the column centre-line for a givenδT is inferred from measurements of testaerosols. An accurate Kohler model is applied to determinethe theoretical supersaturations on which the calibration isbased. For the CCNc’s measurements to be valid it must beassumed that the temperatures in the column are set repeat-edly to the same values for each set pointT , over the courseof an experiment lasting≈2 weeks during which the CCNcis in continuous operation, this appears to be the case. Whatis not assumed is that the centrelineS corresponds directly tothe set pointT s. For the experiments presented here the tem-perature of the sample aerosol is maintained close to 20◦C,this ensures the absoluteT s do not vary significantly whensetting the same supersaturation. All the measurements wereperformed at sea level so pressure variation was minimised.

For both the HTDMA validation and CCNc calibrationtwo test aerosols were used. By using separate standards,bias as a result of factors other than the effective RH andS

such as the DMA’s mis-sizing significantly can be discountedas this would cause different correction factors for the 2 testaerosols.

The derivation ofSc from the CCNc relies on the assump-tions described previously. These may limit the applicabilityof the technique in certain environments. In particular themixing state may affect the derivation ofSc. The simulta-neous HTDMA measurements allow the mixing state to beassessed.

Acknowledgements.Nicholas Good was supported by a NERCstudentship (NER/S/A/2005/13221).

Edited by: S. Buehler

References

Allan, J. D., Baumgardner, D., Raga, G. B., Mayol-Bracero, O.L., Morales-Garcıa, F., Garcıa-Garcıa, F., Montero-Martınez, G.,Borrmann, S., Schneider, J., Mertes, S., Walter, S., Gysel, M.,Dusek, U., Frank, G. P., and Kramer, M.: Clouds and aerosols inPuerto Rico - a new evaluation, Atmos. Chem. Phys., 8, 1293–1309, doi:10.5194/acp-8-1293-2008, 2008.

Alofs, D. J. and Balakumar, P.: Inversion to Obtain Aerosol SizeDistributions from Measurements with a Differential MobilityAnalyzer, J. Aerosol Sci., 13, 513–527, 1982.

Chen, D. R., Pui, D. Y. H., and Kaufman, S. L.: Electrospraying ofConducting Liquids for Monodisperse Aerosol Generation in the4 Nm to 1.8 Mu-M Diameter Range, J. Aerosol Sci., 26, 963–977, 1995.

Cubison, M. J., Coe, H., and Gysel, M.: A modifiedhygroscopic tandem DMA and a data retrieval methodbased on optimal estimation, J. Aerosol Sci., 36, 846–865,doi:10.1016/j.jaerosci.2004.11.009, 2005.

Duplissy, J., Gysel, M., Sjogren, S., Meyer, N., Good, N., Kammer-mann, L., Michaud, V., Weigel, R., Martins dos Santos, S., Gru-ening, C., Villani, P., Laj, P., Sellegri, K., Metzger, A., McFig-gans, G. B., Wehrle, G., Richter, R., Dommen, J., Ristovski, Z.,Baltensperger, U., and Weingartner, E.: Intercomparison studyof six HTDMAs: results and recommendations, Atmos. Meas.Tech., 2, 363–378, doi:10.5194/amt-2-363-2009, 2009.

Fuchs, N. A.: On the Stationary Charge Distribution on AerosolParticles in a Bipolar Ionic Atmosphere, Pure. Appl. Geophys.,56, 185–193, 1963.

Gunthe, S. S., King, S. M., Rose, D., Chen, Q., Roldin, P., Farmer,D. K., Jimenez, J. L., Artaxo, P., Andreae, M. O., Martin, S.T., and Poschl, U.: Cloud condensation nuclei in pristine tropi-cal rainforest air of Amazonia: size-resolved measurements andmodeling of atmospheric aerosol composition and CCN activity,Atmos. Chem. Phys., 9, 7551–7575, doi:10.5194/acp-9-7551-2009, 2009.

Gysel, M., McFiggans, G., and Coe, H.: Inversion of tandem differ-ential mobility analyser (TDMA) measurements, J. Aerosol Sci.,40, 134–151, doi:10.1016/j.jaerosci.2008.07.013, 2009.

Hoppel, W. A., Frick, G. M., and Larson, R. E.: Effect of Non-precipitating Clouds on the Aerosol Size Distribution in the Ma-rine Boundary-Layer, Geophys, Res. Lett., 125–128, 1986.

Irwin, M., Allan, J., and McFiggans, G.: Evaluation of errors inderived data products from water uptake measurements of atmo-spheric particulatesl, Atmos. Meas. Tech. Discus., in press, 2010.

Juranyi, Z., Gysel, M., Duplissy, J., Weingartner, E., Tritscher, T.,Dommen, J., Henning, S., Ziese, M., Kiselev, A., Stratmann, F.,George, I., and Baltensperger, U.: Influence of gas-to-particlepartitioning on the hygroscopic and droplet activation behaviourof alpha-pinene secondary organic aerosol, Phys. Chem. Chem.Phys., 11, 8091–8097, 2009.

Keady, P., Quant, F., and Sem, G.: Differential mobility particlesizer: a new instrument for high-resolution aerosol size distribu-tion measurement below 1µm., TSI Quarterly, 9, 3–11, 1983.

King, S. M., Rosenoern, T., Shilling, J. E., Chen, Q., and Martin,S. T.: Increased cloud activation potential of secondary organicaerosol for atmospheric mass loadings, Atmos. Chem. Phys., 9,2959–2971, doi:10.5194/acp-9-2959-2009, 2009.

Knutson, E. O. and Whitby, K. T.: Aerosol Classification by Elec-tric Mobility: Apparatus, Theory, and Applications, J. Aerosol



Sci., 6, 443–451, 1975.Lance, S., Medina, J., Smith, J. N., and Nenes, A.: Mapping the

operation of the DMT Continuous Flow CCN counter, AerosolSci. Tech., 40, 242–254, 2006.

Liu, B. Y. H., Pui, D. Y. H., Whitby, K. T., Kittelson, D. B.,Kousaka, Y., and McKenzie, R. L.: Aerosol Mobility Chromato-graph - New Detector for Sulfuric-Acid Aerosols, Atmos. Envi-ron., 12, 99–104, 1978.

Maricq, M. M.: On the Electrical Charge of Motor Ve-hicle Exhaust Particles, J. Aerosol Sci., 37, 858–874,doi:10.1016/j.jaerosci.2005.08.003, 2006.

McFiggans, G., Artaxo, P., Baltensperger, U., Coe, H., Facchini, M.C., Feingold, G., Fuzzi, S., Gysel, M., Laaksonen, A., Lohmann,U., Mentel, T. F., Murphy, D. M., O’Dowd, C. D., Snider,J. R., and Weingartner, E.: The effect of physical and chem-ical aerosol properties on warm cloud droplet activation, At-mos. Chem. Phys., 6, 2593–2649, doi:10.5194/acp-6-2593-2006,2006.

Nenes, A., Chuang, P. Y., Flagan, R. C., and Seinfeld, J. H.: Atheoretical analysis of cloud condensation nucleus (CCN) instru-ments, J.Geophys. Res. [Atmos], 106, 3449–3474, 2001.

O’Dowd, C. D., Facchini, M. C., Cavalli, F., Ceburnis, D., Mircea,M., Decesari, S., Fuzzi, S., Yoon, Y. J., and Putaud, J. P.:Biologically-driven organic contribution to marine aerosol, Na-ture, 431, 676–680, doi:10.1038/nature02959, 2004.

Petters, M. D., Prenni, A. J., Kreidenweis, S. M., and DeMott, P. J.:On measuring the critical diameter of cloud condensation nucleiusing mobility selected aerosol, Aerosol Sci. Tech., 41, 907–913,2007.

Rissman, T. A., Varutbangkul, V., Surratt, J. D., Topping, D. O.,McFiggans, G., Flagan, R. C., and Seinfeld, J. H.: Cloud con-densation nucleus (CCN) behavior of organic aerosol particlesgenerated by atomization of water and methanol solutions, At-mos. Chem. Phys., 7, 2949–2971, doi:10.5194/acp-7-2949-2007,2007.

Roberts, G. C. and Nenes, A.: A continuous-flow streamwisethermal-gradient CCN chamber for atmospheric measurements,Aerosol Sci. Technol., 39, 206–221, 2005.

Rose, D., Gunthe, S. S., Mikhailov, E., Frank, G. P., Dusek, U.,Andreae, M. O., and Poschl, U.: Calibration and measurementuncertainties of a continuous-flow cloud condensation nucleicounter (DMT-CCNC): CCN activation of ammonium sulfateand sodium chloride aerosol particles in theory and experiment,Atmos. Chem. Phys., 8, 1153–1179, doi:10.5194/acp-8-1153-2008, 2008.

Sjogren, S., Gysel, M., Weingartner, E., Baltensperger, U., Cubison,M. J., Coe, H., Zardini, A. A., Marcolli, C., Krieger, U. K., andPeter, T.: Hygroscopic growth and water uptake kinetics of two-phase aerosol particles consisting of ammonium sulfate, adipicand humic acid mixtures, J. Aerosol Sci., 38, 157–171, 2007.

Swietlicki, E., Zhou, J., Covert, D. S., Hameri, K., Busch, B.,Vakeva, M., Dusek, U., Berg, O. H., Wiedensohler, A., Aalto, P.,Makela, J., Martinsson, B. G., Papaspiropoulos, G., Mentes, B.,Frank, G., and Stratmann, F.: Hygroscopic properties of aerosolparticles in the north-eastern Atlantic during ACE2, Tellus B, 52,201–227, 2000.

Swietlicki, E., Hansson, H. C., Hameri, K., Svenningsson, B.,Massling, A., McFiggans, G., McMurry, P. H., Petaja, T.,Tunved, P., Gysel, M., Topping, D., Weingartner, E., Bal-tensperger, U., Rissler, J., Wiedensohler, A., and Kulmala, M.:Hygroscopic properties of submicrometer atmospheric aerosolparticles measured with H-TDMA instruments in various envi-ronments – a review, Tellus B, 60, 432–469, 2008.

Topping, D. O., McFiggans, G. B., and Coe, H.: A curved multi-component aerosol hygroscopicity model framework: Part 1- Inorganic compounds, Atmos. Chem. Phys., 5, 1205–1222,doi:10.5194/acp-5-1205-2005, 2005.

Topping, D. O., McFiggans, G. B., and Coe, H.: A curved multi-component aerosol hygroscopicity model framework: Part 2 -Including organic compounds, Atmos. Chem. Phys., 5, 1223–1242, doi:10.5194/acp-5-1223-2005, 2005.

Vana, M., Tamm, E., Horrak, U., Mirme, A., Tammet, H., Laakso,L., Aalto, P. P., and Kulmala, M.: Charging state of atmosphericnanoparticles during the nucleation burst events, Atmos. Res.,82, 536–546, 2006.

Wiedensohler, A.: An Approximation of the Bipolar Charge-Distribution for Particles in the Sub-Micron Size Range, J.Aerosol Sci., 19, 387–389, doi:10.1016/0021-8502(88)90278-9,1988.

Williams, P. I., McFiggans, G., and Gallagher, M. W.: Latitu-dinal aerosol size distribution variation in the Eastern AtlanticOcean measured aboard the FS-Polarstern, Atmos. Chem. Phys.,7, 2563–2573, doi:10.5194/acp-7-2563-2007, 2007.


Instrumentational operation and analytical methodology for ...

Documents

Transcript of Instrumentational operation and analytical methodology for ...